Separating solids from suspensions, optionally combined with washing of the solids in hydraulic washing columns is a well known technique for concentration of food products (coffee extract, juices), for recovering crystals in pure form from suspensions, for desalination of sea water, and the like.
An apparatus for concentrating a suspension is known from U.S. Pat. No. 2,854,494. This patent describes a crystal wash column in which in a cylindrical room a crystal suspension is concentrated by directing this suspension along a section of the cylindrical wall which contains one or more filters. At least a substantial portion of the liquid in the suspension is removed through these filters. The crystals substantially rid of liquid, thus in a concentrated suspension, are transferred along the filters and removed by a way different from that of the liquid. In this known apparatus the liquid and the crystals of the suspension to be treated initially move in the same direction. Via the filters mounted in the wall, however, the greater part of the liquid is removed sideways, while the crystals continue to move in the same direction.
A disadvantage of this hydraulic washing column resides therein, that it is very difficult to develop a column based on this principle suitable for large scale commercial operation. More in particular, with column sizes of over about 75 cm diameter, problems may occur with respect to pressure build-up and homogeneity of the wash/separation front, depending, among others, on the nature of the solid material and/or the particle size and shape thereof.
More in particular in the neighbourhood of the filter a densely packed crystal bed is formed, on which a big force has to be exercised for its transfer due to the fact that the friction between such a packed bed and the wall of the column can become very considerable with larger column diameters, due to the increasing radial pressure gradient. In other words, the ratio between friction forces and transport forces increases. This big force exercised on the concentrated suspension involves the compression of the packed bed, which can result in deforming or damaging the crystals. The consequence thereof is a further increase in pressure drop and a decrease in throughput capacity of the column. Further an increased energy consumption may be the consequence too.
A further disadvantage of this known apparatus is that a homogeneous flow of washing liquid through the crystal bed becomes more difficult with increasing column diameter. Since the filters are located in the column wall in the neighbourhood of the filters there is not an evenly distributed axial liquid flow, but an increased radial liquid flow directed towards the filter. This disturbing effect of the filters mounted in the wall on the liquid flow in the crystal bed increases with a larger column diameter and results in an uneven profile of the washing front.
In U.S. Pat. No. 4,735,781 a further development of this known column is described, which development consists of the provision of a number of filter tubes inside the treatment zone.
In a column of the type described in U.S. Pat. No. 4,735,781 extrapolation of the literature data shows, that typically about 300 tubes of 2 cm would be required for a 110 cm diameter column.
It is well known, that the presence of tubes in a packed bed of solids leads to a disturbance thereof. As a result, bridge formation and channelling can occur in the bed. Also the increased wall surface area leads to increased friction forces, as discussed above.
Further problems encountered with this construction are that a homogeneous distribution of the slurry over the entire cross-section of the column is difficult, leading to that it is difficult to create a homogeneous bed, and that is it difficult to perform an accurate measurement of the bed height.
Finally it is to be noted the tubes require very narrow dimensional tolerance and precise positioning, which aspects, together with the increased amounts of materials, make the column more expensive to construct.
Quite likely those are the reasons that up to now this column has never been demonstrated successfully on an industrial scale.